Upload
others
View
10
Download
0
Embed Size (px)
Citation preview
ED 302 087
AUTHORTITLE
DOCUMENT RESUME
FL 017 731
Teschner, Richard V."Provided There Is an Adequate Exposure to This L2 inthe School and Environment and Sufficient Motivationto Learn It": The Applicability (at times "pace"Cummins): of Majority-Language Immersion Programs toLimited- or Non-English-Proficient Spanish-L1Grade-School Populations in the United States-MexicanBorder-Area.
PUB DATE Oct 88NOTE 39p.; Revised version of a paper presented at the
Annual Meeting of the Linguistic Association of theSouthwest (October 1988).
PUB TYPE Information Analyses (070) -- Reports -Evaluative /Feasibility (142) -- Speeches/ConferencePapers (150)
EDRS PRICE MF01/PCO2 Plus Postage.DESCRIPTORS *Bilingual Education Programs; Comparative Analysis;
Educational Strategies; Elementary SecondaryEducation; *English (Second Language); *ImmersionPrograms; Langlage of Instruction; LearningMotivation; Limited English Speaking; Non EnglishSpeaking; *Program Effectiveness; Second LanguageLearning; *Spanish; Teaching Methods; TransitionalPrograms
IDENTIFIERS Cummins (James); *El Paso Independent Schdol DistrictTX; Whole Language Approach
ABSTRACTThe El Paso (Texas) Independent School District's
district -wide program-in majority-language (English) immersion isexamined to determine_the source of its success. The program'ssimilarities to and differences from the Canadian immersion model areexplored. The program's superior results in comparison with the samedistrict's transitional bilingual education programs are-looked at interms of James Cummins' discussions of bilingual education. It isconcluded that the immersion program's superiority over thetransitional programs can be explained largely in terms of itsmaximal use of English and its whole language approach. The bilingualimmersion program approach is recommended not as a panacea, but as asuperior educational alternative. A list of 79 references-isincluded. (MSE)
***********************************************************************
* Reproductions supplied by EDRS are the best that can be made ** from the original document. *
***********************************************************************
4
N. .CJ . "'Provided There Is An Adequate Exposure to This L2 in the School and Environment
(:).
lmj and Sufficient Motivation to Learn It': The Applicability (at times pace Cummins)C:1ref of Majority-Language Immersion Programs to Limited- or Nan-English-Proficient
Spanish -L1 Grade-School Populations in the United States-Mexican Border Area."*tilliVeRMISSION TO REPRODUCE THIS
MATERIAL HAS BEEN GRANTED BY
T- eAd0. 2A Richard V. Teschner
TO THE EDUCATIONAL RESOURCESINFORMATION CENTER (ERIC)."
University of Texas at El Paso
U.S. DEPARTMENTCF EDUCATOROffice of Educational ResearCh and Improvement
EDUCATIONAL RESOURCES INFORMATIONCENTER IERIC)
Thrs document has been reproduced asreceived from the person or organizationoriginating it.
O Minor changes have been made to improvereproduction quality.
Points of vieWOrOpiniOne stated in this docu-ment do not necessanly represent officialOERI posdion or PoliCY.
What most transitional bilingual education programs have in common is this:
non- or limited-English proficient students are taught English as a Second lang-
uage (ESL) during one class period or more on any given day, and are taught most
or all their "content" subjects through the medium of the home language. This
is what is broadly mandated in the various pieces of legislation both federal
and state to be approved since the passing by Congress in 1968 of the Bilingual
Education Act. Typical of state rules are those of Texas, as cited in the Texas
Education Code (1982, 21.454, "Program Content: Method of Instruction," p. 164):
to be established is "a full-time program of dual-language instruction that pro-
vides for learning basic skills in the primary language of the students of lim-
ited English proficiency who are enrolled in the program, and that provides for
carefully structured and sequenced mastery of English language skills." Because
of the limited flexibility--or perhaps outright contradition--set forth in the
Texas Education Agency's (1982) Comprehensive Instruction guidelines (section
77.353, "Program Content; Method of Instruction," p. 453), it is not possible
to speak of curricular uniformity in bilingual education programs in Texas let
tl alone throughout the United States; thus according to the Texas Education Agency,
Obilingual education shall develop basic skills of comprehending, speaking, read-
An1I ing and writing in the student's primary language as well as in English, and
*This article is a revised version of the Presidential Address presented at the
annual meeting of LASSO {Linguistic Association of the Southwest}, October, 1988.
2
2
"subject matter and concepts" shall be taught in the student's primary language
as well as in English.
Enter. Canadian immersion education (henceforth CIE;, born, as is well
known, in 1967 in the Montreal suburb of St. Lambert, whose anglophone parents
were disturbed by the long-time failure of French FLES (Foreign Language in the
Elementary School) to produce French-proficient offspring, and who in any event
saw the proverbial handwriting on the wall, realizing that Quebec was rapidly
headed toward de jure French monolingualism, and that without French, future
generations could not earn a living in the province. CIE operates thus: All
children in the program's first year (which can be kindergarten or later) arrive
as monolinguals in the same home language (English in this instance). The teach-
er, who is bilingual though natively francophone, uses only the target language
(French) for all classroom purposes both instructional and non-instructional,
yet the children are allowed--indeed expected--to address the teacher in the
home language until they are ready for spontaneous, self-generated oral produc-
tion of the L2, which typically occurs in about a year. Aural comprehension
occurs sooner--within three to four months--and, in fact, begins to occur frcnn
the very beginning, in that initial language use is highly contextualized, and
can involve physical responses to teacher-generated requests.
At the heart of immersion-education is the use of the target language as
both medium of instruction and instructional text. At some point--for the St.
Lambert students it was at the end of the second grade, i.e., the students'
third year in the program--the home language is introduced as subject material.
The St. Lambert experimentals were subjected from the outset to a wide battery
of tests, and were compared on multiple measures with both monolingual anglo-
phone youngsters following the same school district's English-medium curriculum
and with monolingual francophone students following an analogous curriculum in
an adjacent district. To sum up the St. Lambert program's voluminous empirical
findings (see Lambert and Tucker 1972), replicated subsequently by scores of
3
districts throughout Canada from the early seventies onward, the immersion
experimentals acquired French far more successfully than their older siblings
ever had done through FLES; showed no deficit in the academic content areas (in
particular math); and--despite initially depressed performance--showed no long-
range deficit in their ability to read and write English. In sum, CIE repre-
sented a success story, at least through about grade five, when researchers
such as Spilka (1976) began to note a leveling off in the students' progress
in French, typified by fossilizing interlanguage--the so-called Plateau Effect,
which subsequent work by Hammerly (1987) and others has confirmed remains oper-
ant all the way thrc ih the twelfth grade. The Plateau Effect would appear to
result from the ethnolinguistically-segregated nature of Canadian society:
though CIE students in those parts of Canada where anglophones potentially en-
joy access to sizeable natively-francophone communities were being educated via
French, the long-standing animosity between "the French" and "the English" in
Canada (see for example Wardhaugh 1982) prevented English-Canadian parents from
taking the final, logical step of enrolling their children in French-Canadian
schools so that peer influence could take effect, eroding the interlanguage mold
and giving rise to fully authentic Canadian French. But such was not to be. In
any event, Plateau Effect or no, the French the immersion students speak is con-
siderably better than the French the previous generations hardly ever learned.
Around the 'middle 1970s, when word of Canadian successes started filtering
south, American educators unconvinced by or opposed to what I have already des-
cribed as transitional bilingual education began to question whether the Canadian
model could be applied here. In his voluminous writings on bilingual education
and allied topics (see REFERENCES), the Irish-Canadian psychologist Jiml Cummins
for the most part argues--especially in his earlier writings--against the possi-
bility or the desirability of applying the Canadian immersion model to American
classrooms, serving limited- or non-English-proficient students, whom Cummins
terms language minorities. Thus in Cummins 1978c:855 he states that while "Im-
4
-z"
4
mersion programs for majority language children have proved extremely success-
ful, resulting in high levels of second language (L2) skills at no cost to first
language (L1) skills," the "home-school language switch has been shown to result
in low levels of achievement in both Ll and L2 in minority language children."
In a discussion of his well-known "developmental interdependence" and "threshold"
hypotheses, Cummins (1979c:243) insists that "for the child whose input conceptual-
linguistic knowledge is not conducive to the development of literacy skills,"
the hypotheses imply that "initial instruction should be through the medium of
Ll." While Cummins admits subsequently (1981f:44) the existence of a variable--
ambiental exposure to target language--whose ramifications he deals with else-
where and which will be discussed later in the present paper, he continues to
insist that any increased exposure to L2 must be postponed ("this increased ex-
posure should not come in the early grades where the instructional emphasis
should be on Ll in order to develop the. conceptual apparatus required to make
English context-reduced input comprehensibTe").
However as Cummins' own work evolved, and research findings from various
programs were made manifest, Cummins increasingly admitted the possibility that
some features of Canadian-style immersion education might be practicable in
American language-minority school settings. A neutral, take-no-sides report
on several types of U.S. immersion programs for non-anglomonomatriphones appears
in Cummins 1983a:379-380. One year later (1984b:156), Cummins does not exclude
adaptations of the Canadian model when he writes that "In light of the analysis
of minority student underachievement presented in previous chapters, there ap-
pears to be little reason why such a 'genuine immersion' programme might not be
successful. However, the relevance of sociocultural variables associated with
student ambivalence vis-a-vis Ll and L2 suggests that there should be an Ll
component from the beginning of the programme." (Themes such as language and
cultural ambivalence along with increased emphasis on sociocultural variables
are characteristic of Cumilins' later wo;k and will be discussed below.) The
5
5
same source offers up what could be termed a qualified endorsement of both
majority and minority language immersion programs of the sort that a very
small number of American school districts had cautiously begun to promote.
Sounding a note of warning about transitional bilingual education programs
that he develops elsewhere into a veritable symphony, Cummins writes: "That
there be genuine sustained reinforcement of minority students' Ll (as opposed
to its transitional use in most U.S. bilingual programmes) is probably ultimately
more important than the specific amount of Ll instructional time (e.g., 80% in
San Diego early grades, 20% in McAllen) or the language in which reading is
introduced. The relative efficacy of majority language and minority language
bilingual programmes is an empirical question at this point." (1984b:160) As
the quote suggests, a majority-language bilingual (immersion) program is one
which gives greater emphlsis to the languaye of the national majority population
(here English), while a minority-language bilingual program emphasizes the minor-
ity group's home and community language (here Spanish).
It is the purpose of the present paper to examine thoroughly a district-wide
bilingual program in majority-language immersion, that of the El Paso Independent
School District (henceforth EPISD); pinpoint that program's defects as well as
its successes; explain why that program has so far been achieving superior re-
sults when compared to those of the control group's program; and attempt to ex-
plain these results in terms of Cummins' discussions of bilingual education. At
present it is not possibile to respond to Cummins' suggestion that the relative
efficacy of majority versus minority language bilingual programs serving similar
populations be discussed, since no minority-language bilingual immersion program
has been subjected to any significant-empirical examination to date. 2 Instead,
the EPISD's bilingual program in majority-language immersion will be compared
with the same district's transitional bilingual education program.
What precisely is bilingual (majority-language) immersion, and how does it
differ from both the Canadian-style immersion education as well as transitional
6
bilingual education described above? Answers to these questions along with
the data presented here on the EPISD's bilingual immersion program (henceforth
BIP) derive from the EPISD's Office for Research and Evaluation's Interim Report
of the Five-Year Bilingual Education Pilot, 1986-87 SChool Year (henceforth
Interim Report), issued in-house July 1987 in response to a request from the
Texas Education Agency (TEA).3
Transitional bilingual education (henceforth TBE) allows for the possibil-
ity of teachers presenting content areas via the medium of the target language,
but in practice (or at least in the EPISD) does not do so; instead, English is
a subject matter only, taught for 60 minutes a day in grade one and 90 minutes
a day subsequently; all other subjects are taught in Spanish. By contrast, the
bilingual immersion program--BIP--uses mainly English as the medium of instruc-
tion for content subjects but makes a point of using what the Interim Report
calls "comprehensible English," i.e., a context-embeddedl code, abounding in
concrete referentiality. Students are not discouraged from responding in Span-
ish. though English responses are encouraged when the language of transmission
is English. BIP does use some Spanish as a medium of instruction for content
subjects from the very outset, and in that sense BIP differs from Canadian im-
mersion education (CIE), where only the target language is used for that purpose
initially. As is true of CIE, students are given daily instruction in Spanish
language arts (reading and writing); the major difference between the two prog-
rams is that BIP offers Spanish language arts from the very outset, while CIE
postpones Ll language arts until the end of grade two.
The Interim Report stresses that BIP should not be viewed as some sort of
flip side of the TBE instructional coin; indeed, one is struck by the extent to
which the difference between BIP and TBE is generational: TBE, conceived in
the 1960s.and first put into effect in El Paso around 1970 (though not expanded
district-wide until 1977), is clearly the intellectual offspring of Spanish-
English contrastive linguistics, a state-mandated course required since about
7
1970 of every candidate for bilingual certification in Texas. As the Interim
Report notes, TBE focuses the child's attention contrastively on the details of
language, i.e., phonetics and grammar rules, while BIP is inspired by that the
report calls the Whole Language Approach, a reciprocal interactive system. ,
BIP consciously seeks to contextualize both oral and written language and to
use language in the classroom for authentic communicative purposes, or at least
for purposes as authentically communicative as a classroom setting allows.
There are three BIP components: English Language Arts (analogous in purpose
if not methodology to the ESL component of TBE), Native Language Cognitive
Development (where Spanish is consistently the medium and often the subject of
instruction), and science, social studies and math as "sheltered content areas."
("Sheltered" means English-medium input that is deliberately reduced in complex-
ity and abstractness, especially at the beginning; see for example Cummins
1979b). English Language Arts is taught for either 90 or 120 minutes per day,
depending on the level of the students; the Interim Report states that the
methods employed are Natural Approach, Total Physical Response, and other com-
municative orientations. Reading and writing are fostered through such inter-
active process strategies as Storytelling, Daily Journal Writing, and Group and
Individual Publishing. Native Language Cognitive Development is offered for
from 60 to 90 minutes per day; according to the Interim Report (1987:10), the
"objective of this component is to develop concepts, literacy, cognition and
critical thinking skills in Spanish. It is during this period that instruction
and student-teacher interaction are entirely in Spanish. The more demanding
content area concepts may also be introduced during Native Language Cognitive
Development."
How well do BIP and TBE students test? Since Chapter 77 of the Texas Ad-
ministrative Code (1982) calls for annual testing of all- students identified as
limited English proficient, including all those enrolled in the EPISD in either
TBE or BIP, the Interim Report was able to present data from over a thousand
individuals, all of whom took the reading and language arts sections of the Iowa
Tests of Basic Skills and the Oral Language Dominance Measure as well as "TEAMS,"
the Texas Educational Assessment of Minimum Skills test. My reportage of test
results will focus on grade three (the highest reached by BIP at time of test-
ing), in part because not all grade one (and sometimes not all grade two) students
are tested in all subjects for a variety of reasons. (Thus state law allows stu-
dents whose language is so limited that they cannot take the test to be exempted
from it.) In any event, it is the grade three scores that will be of greatest
interest to us. Before reporting any test results I should make it clear that
while I am in sympathy and essential agreement with the conclusions reached by
my antepenultimate predecessor in this office, Carole Edelsky, in her 1985 LASSO
presidential address "The Effect of 'Theory' on Theory--And Other. Phenomena"
(published in revised form as Edelsky 1984, and expanded upon as Edelsky et al.
1983), I admit I cannot argue with Cummins and Swain (1983:27) in their response
to Edelsky et al. 1983 when they (C & S) state: " . . . we do not believe that
the real world is sufficiently clear cut to dismiss all uses of standardized
tests, under all conditions, for all students as 'irrelevant nonsense'." Cummins
and Swain aptly point out that " . . . standardized tests can provide useful in-
formation under some conditions and for some students. For example, the spread
of French immersion programs across Canada is due in no small measure to the
repeated finding that students did not suffer {a decline in} English academic
skills as measured by standardized norm-referenced tests." (Cummins and Swain
1983:28-29)
For those grade three students tested district-wide in March of 1987, the
normal-curve-equivalent-based mean scores were higher for BIP students than
for TBE students in all three subject areas of the English-medium Iowa Test
(reading, language and math), though the BIP advantage was statistically sig-
nificant only for reading and language, with the math means too close to call.
It should be noted that both BIP and TBE scores were lowest in reading and
9
9
highest in language. The BIP mean scores exactly equalled the national average
and the TBE scores fell five points below it.
The EPISD uses the Oral Language Dominance Measure as its oral language
testing instrument in grades one through three; this test measures spoken
language proficiency in both English and Spanish, and its scale ranges from
a low of one to a high of five. It is on the English proficiency component
that BIP students unequivocally outperform their TBE peers: in grade three,
74% of BIP students but only 56% of TBE students achieved the high score of
five, and only 18% of BIP students versus 29% of TBEs scored between one and
three, i.e., low. What about spoken language proficiency in Spanish? In grade
one, 84% of BIP and 88% of TBE students achieved the highest scores, in grade
two 96% and 98% respectively, and in grade three 97% of both cohorts did so.
There is little if any difference then between BIP and TBE when it comes to
maintaining and developing oral proficiency in Spanish; theTBEs may have a
slightly higher score at first but the two groups perform at an equal--and an
equally high--level by grade three.
Since 1979, school districts in Texas have been required by law to parti-
cipate in a State-developed minimum basic skills testing program, whose current
version is the TEAMS test referred to earlier. These are criterion-referenced
tests, i.e., measurement instruments that relate test items to specific learning
objectives or levels of proficiency in skills which students have been taught.
EPISD students take TEAMS in grades 1, 3 and 5--this last not reported here
since the BIP program had only entered its third year as of March 1987. Span-
ish versions of TEAMS for grades one and three were administered for the first
time in that year. EPISD Language Proficiency Assessment Committees decided
which students would take the Spanish version and which the English version.
In grade three, 87% of the BIP students (n = 535) and 66% of the TBE students
(n = 529) took the English TEAMS while only 13% BIP students (80) but 34%
TBE students (268) took the Spanish teams. While the low number--both absolute
10
and relative--of BIP students taking the Spanish TEAMS may vitiate the validity
of the comparisons, I will go ahead and present them anyway if only for their
heuristic value.
Both groups did equally well on the English version of the TEAMS math
test (80% mean mastery), and almost as well as the district mean (85%). While
the BIP students' English reading scores (47% mastery) exceeded TBE students'
(39%), and at statistical significance, the disparity between these levels of
mastery and the mean district-wide non-bilingual-program-mastery level- -70 % --
bears thinking about. Similar concern can be shOwn about English writing scores:
BIP 44% mastery, TBE 40% mastery, district-wide 65% mastery. On the Spanish ver-
sion of the math test, the tables are turned (though again, bear in mind that
the Spanish-version BIP cohort is small): BIP 74% mastery, TBE 87% mastery,
state-wide bilingualprogram 79% mastery. (We can assume that "state-wide
bilingual program" means "TBE elsewhere in Texas.") Thus in math in Spanish,
the 80 BIP students undershot the state-wide mark by five percentage points
and the 268 TBE students exceeded it by eight. In Spanish reading, both El
Paso groups outperformed state means: TBE 97% mastery, BIP 89% mastery, and
86% state-wide mastery. Similar results are manifest in Spanish writing: 95%
TBE mastery, 91% BIP mastery, and 86% state-wide mastery.
In general terms, the results from these three batteries of tests provide
very suggettive evidence supporting several conclusions.
Conclusion number one appears commonsensical, and is not entirely gainsaid
by Cummins despite his repeated opposition to what he terms the fallacy of the
"maximum exposure assumption" (thus Cummins 1980c:51, 1984c:33, 1984:264-265):
the more you are exposed to meaningful input in a language, the better you will
do in it. BIP students did better--though in just :me component to any statis-
tical significance--on the Iowa tests, which are administered via English; un-
equivocally outperformed TBE students on the English portion of the Oral Language
Dominance Measure; and did better at English writing and especially English
11
reading as measured by, the TEAMS test. Only in the TEAMS test's English-version
math exam did both groups do equally well on an English-medium test. But as the
Interim Report (1987:61) notes, "BIP's success in mathematics is interesting in
light of a major argument for {transitional bilingual education} : that content
area instruction must be carried out in Spanish to ensure continuation of {con-
tent area} learning while English is being learned." Of course the argument
works both ways: the fact that TBE students did as well as their BIP kers on
an English-medium math exam has demonstrated that in math at least, knowledge
acquired via the home language readily transfers into the target language. BIP
students, then, are ahead of their TBE peers in English by whatever measurement,
and have not fallen behind them in a crucial content area such as mathematics.
Conclusion number two is not surprising in light of what we have long known
from the research on the products of Canadian immersion education: that BIP
students' performance in Ll language arts (here Spanish) did not differ greatly
from the performance of their TBE counterpa %s who, save for those 60 or 90
minutes a day of ESL, have been educated solely via the medium of the L2. By
the third grade, thp two groups' scores in Spanish oral proficiency are iden-
tical. Recall, however, that there is one important difference between Canadian-
style immersion and El Paso-style BIP: that BIP students are given significant
amounts of Spanish language arts from the very beginning of the program, and
note that even this coursework did not enable BIP students to fully equal--let
alone surpass--the TBE students on the Spanish-medium reading and writLa sec-
tions of tile TEAMS test; though the BIP scores were close, TBE scores lead. So
if the goal of bilingual education is to prOuucl students whose bilingualism is
fully balaticA in all four skills, then BIP is not quite adequate, and a cer-
tain amount of additional work both in and via Spanish may be needed.
The above conclusions constitute a corallary to Cummins in that they demons-
trate the efficacy of an L1- inclusive majority-language immersion program in a
setting such as El Paso, as Cummins in his later writings hypothesized might be
12
12-
possible among language-minority students. Of course El Paso's language min-
ority may not be what Cummins had in mind when he wrote of language minorities,
since in El Paso, nearly 70% of the population is ethnically Hispanic, and per-
haps an equal percentage,of El Paso's Hispanics are Spanish- retentive and/or
-preferent, whether as Spanish monolinguals, Spanish dominants, or balanced
bilinguals. At any given moment, then, at least as much Spanish as English is
being spoken throughout El Paso. Even more to the point, most of those neighbor-
hoods whose schools offer BIP or TBE programs are Spanish-retentive, some strongly
so. In a wider, national context, of course, Spanish is a minority language even
in places like El Paso, yet from a local vantage point it is a very special kind
of minority language.
It beart noting--and here we leave Cummins temporarily--that the goal of
bilingual education in Texas at present is petfectly clear: to mainstream pupils
into English-medium coursework. On this point, the laws of Texas are upon us;
thus I quote from the Texas Education Cod:. 121.45l, Subchapter L, Bilingual
Education and Special Language Programs, "State Policy"):
Public schools are responsible for providing full opportunity for all
students to. become competent in speaking, reading, writing, and com-
prehending the English language. ... The legislature recognizes that
the mastery of basic English language skills is a prerequisite for
effective participation in the state's educational program. The
legislature believes that bilingual education and special language
programs can meet the needs of these students and facilitate their
integration into the regular school curriculum. ... Bilingual edu-
cation or special language programs as defined by this Act shall be
taught {sic} in the public schools only for the purpose of assisting
the learning ability of limited English proficiency students aJd to
enhance the English language.
And there is more, elsewhere, along these same lines. Barring a radical change
13
13
in these laws - -and I foresee none, at least not in the next three or four
decades--the question perforcedly becomes: which of the EPISD's two programs--
BIP or TBE--does the better job of preparing students for that all-important
transition into the all-English schoolday? The El Paso interim data would
indicate that BIP does so. BIP students make more progress in all facets of
English in three years than TBE students--and equal progress in content areas-r
so BIP students will do better in the English-language classroom. While Cum-
mins has argued forcefully (1980a:96 and elsewhere) that since research sug-
gests that "'equality' of academic potential and performande is not attained
{by language-mincrity students} until the later grades of elementary school
. . . {therefore} the full benefit of bilingual instruction may not become ap-
parent until the fifth or sixth year of instruction," it is nontheless counter-
intuitive to assume that TBE students, demonstrably behind in English at the
end of grade three, will have achieved parity with the products of BIP by
grade five. Such at least is my prediction; suffice-it to say that the results
of the final EPISD report, due to be issued the summer of 1989 following the
completion of grade five by both cohorts, are awaited with considerable interest.
That BIP does a better job than TBE is a position with which classroom
teachers themselves agree. The EPISD distributed questionnaires to every BIP
and TBE teaeler in the appropriate grades throughout the district. Rate of
response was remarkably high, as 280 TBE teachers (84%) and 231 BIP teachers
(86%) returned filled -out questionnaires. In general, BIP teachers viewed
their programs much more positively than did TBE teachers theirs. (It should
be noted that almost without exception, teachers in the EPISD's bilingual
programs are bilingual native hispanophones themselves.) The questionnaire
contained ten statements--two modified to reflect the two programs' differing
realities, one deleted from the BIP questionnaire altogether--with which the
respondents were to agree or disagree on a six-point scale (though for the
purpose of tabulation, "disagree" ratings one and two and "agree" ratings
14
14
five and six were colapsed into single categories). Thus 60% of BIP versus
only 40% of TBE instructors agreed that "Most students will succeed in the
regular program after they complete" the program now enrolled in, and 20%
of TBE versus only 7% of BIP instructors disagreed with that statement. Remark-
able discrepancy exists between the two groups' reactions to -a statement about
whether each one's program "notivates students to learn English": 75% of the
BIP versus only 36% of the TBE faculty agreed theirs does, and 32% of the TBE
instructors felt that their program does not so motivate. A highly analogous
statement--the program "successfully develops students' oral English skills"--
drew nearly identical responses: 74% BIP vs. only 37% of the TBE faculty agreed,
and a substantial 28% of the TBE faculty (vs. only 5% of the BIP faculty) dis-
agreed. The two groups were almost equal in their response to the statement
"The program develops and maintains students' Spanish language skills" (59%
TBE and 60% BIP agreement); only small percentages--7% TBE, 4% BIP faculty- -
disagreed with this statement. Similar near-equality reigned in faculty res-
ponses to the following: the program "is successful in teaching children to
read in" Spanish (if TBE) or English (if BIP). Since the BIP program's child-
ren read in English from the very outset, only TBE instructors were asked to
respond to the following statement: "Most students in the program successfully
transfer to reading in English"; 36% agreed, 22% felt they did not, and the
plurality (42%) neither agreed nor disagreed. Substantive differences were
also revealed in responses to three more statements: "This program motivates
students to read and enjoy stories" (BIP 79% agreement and 3% disagreement,
TBE 57% agreement and 14% disagreement); this program "successfully develops
students' grammar, punctuation and spelling skills" (BIP 65% agreement and 11%
disagreement, TBE 44% agreement and 22% disagreement); and this program "en-
courages students to positively identify with their cultural heritage" (BIP
64% agreement and 7% disagreement, TBE 63% agreemenc and 13% disagreement).
The lower level of teacher enthusiasm for TBE as manifested in responses to
15
this last statement is especially striking, since TBE's all-Spanish-in-the-
content-areas curriculum was clearly intended to produce precisely the effect
stated: positive identification with the home language's cultural heritage.
Equally striking is TBE faculty response to the last of the ten statements
("Students in the program benefit from being taught the content areas in Span-
ish"): a very bare majority (51%) agreed, while slightly more than one ot.t of
five (21%) disagreed, and 28% neither agreed nor disagreed. It should be noted
at this point that the overwhelming majority of the teachers in the EPISD's bi-
lingual programs received their training in the University of Texas at El Paso's
College of Education, all of whose bilingual education faculty have long supported
transitional bilingual education with vocal enthusiasm, and that if there is one
concept which is central to the core of that program, it is that students bene-
fit from being taught the content areas in Spanish. (By way of contrast, 83%
of BIP instructors agreed that students in their program "benefit from being
taught the content areas in English" and only 3% disagreed.)
None of these findings concerning the superior performance of BIP students
on standardized tests and the teachers' perceptions of BIP as superior- to TBE
should be surprising to the reader of Cummins, since by 1980 Cummins had reached
the conclusion that "the psychoeducational assumptions underlying transitional
programs are largely invalid." (1980a:83)- Cummins bases this assertion on two
considerations: the invalidity of what Cummins terms the "linguistic mismatch"
hypothesis, and the strong likelihood that "the educational benefits of bi-
lingual education may be cumulative, and thus, aborting the program at an
early stage is unlikely to realize these benefits." (ibid.) The second of
these considerations will not be dealt with here, all the more so because none
of the students from either program had been exited from it, at time of testing
(and would not, indeed, be exited until the end of the fifth grade). The first
consideration--the linguistic mismatch hypothesis--is based on the seemingly
logical assumption that children will only learn well if they are instructed
16
16
via the language they already know well, i.e., 11, an assumption given consid-
erable impetus by the oft-cited 1953 UNESCO report (UNESCO 1953). Cummins
argues that considerable research findings have invalidated linguistic mis-
match as a general tenet; prominent among the findings are the conclusions
drawn from Canada's immersion programs: their products have acquired (along
with French) a course-content mastery that invariably equals or exceeds that
achieved by the student in the English-medium control group. Thus linguistic
mismatch i.e. home-school language switch alone can no longer be used as the
prime let alone the sole justification of bilingual education.
While the conclusion to be drawn from my analysis so far is that the super-
iority of BIP over TBE can be explained largely in terms of BIP's maximal use
of English versus TBE's minimal use of it, one cannot overlook the strong like-
lihood that the context of English exposure in BIP--the reciprocal interactive
"Whole Language Approach" referred to earlier--is at least as important a
Predictive factor as is the fact of greater exposure itself. By 1983, Cummins
had begun to speak frequently of the importance of comprehensible input ("Lang-
uage acquisition is largely dependent on students receiving sufficient compre-
hensible input in the target language" (1983a:373)) and to cite the works of
Stephen Krashen often. Underlying the principle of comprehensible input "is
the importance of meaningful communication. When this central language func-
tion is ignored in classroom instruction, learning is likely to be by rote and
supported only by extrinsic motivation." (Cummins 1983a:377) Cummins speaks
directly to the superiority of programs such as BIP when he notes that the
contrast between "immersion and traditional second-language programs can be
interpreted within the framework of the comprehensible input principle -in that
communication between teacher and students in the immersiTfirogram is embedded
in a meaningful concrete context and supported by a wide range of paralinguistic
cues which allow students to. infer the intended meaning and simultaneously
acquire the second language. This, however, is not the case in traditional
17
second-language programs which tend to emphasize language drills in isolation
from authentic communication." (1983a:378) Reference to the importance of
reciprocal interactive L2 facilitation is found elsewhere throughout Cummins'
more recent writings (1981f:14, 1983b:112-113 and 125, 1984b:114 and 224-225,
1984f:67). Only by factoring out one of the two probable determinants of BIP's
success--on the one hand the higher degree of exposure to English, on the other
hand BIP's use of a reciprocal interactive approach to learning versus TBE's
more traditional pedagogy--could this issue be clarified. In 1983 Cuiddins
noted that "virtually no research information is available on the effects of
monolingual immersion for minority students" (19834i379), and this is precisely
the type of information we would need in order to determine which aspect of BIP
accounted for its success or whether both aspects of the program were equally
responsible for the superiority of its results.
I have already noted that the overwhelming reality of American bilingual
education is its intent to serve as a transition to "mainstream" education via
English. At least by the time the initially hispanomonomatriphonic youngster
who began a bilingual program in kindergarten or first grade reaches grade
seven (and often before), he or she is expected to have mastered enough English
to handle a program whose content areas are taught solely in that language.
While recalling that at least in the EPISD the product of whichever bilingual
program--BIP or TBE--is exempted from entering the mainstream classroom until
grade seven if necessary if his or her command of English is still judged in-
adequate, I am nonetheless bothered by the scores achieved by BIP and TBE
students alike on the grade three TEAMS tests of English reading and writing.
It will be recalled that while BIPs achieved a 47 percentile and thereby
did better than TBEs, who only achieved a reading score of 39, the district
as a whole, i.e., those 3,860 students participating in the English version of
the TEAMS reading test in grade three, achieved a 70 percentile average. If
We break down the reading test score into its component parts, we note that
18
18
the district-wide score particularly outdistances the BIP and the TBE scores in
the areas of "sight words," "context clues" and "predicting outcomes," that is,
in the realm of general comprehension as opposed to mechanics ("phonics," "table
of contents" and the like). Scores from the grade three. TEAMS writing test
(English version) are similarly disturbing. The BIP advantage almost disappears
(44th percentile as opposed to a TBE 40th percentile), and neither group even
approaches the district-wide percentile, which is 65. Again, the differences
are at their most evident in global activities; thus on "passing score on para-
graph," i.e., a free composition exercise which required students to generate
original language rather than manipulate prepared language, BIP students scored
54, TBE students 50, and district-wide 75; differences between the three groups
on "proofreading," "spelling," "punctuation," "capitalization" and the like
were minimal by comparison.
It is not for nothing that BIP is an immersion program. Like the products
of Canadian immersion in French, El Paso hispanomonomatriphones in English BIP
score noticeably better in English and on tests using English than their "Eng-
lish FLES" TBE counterparts, but not as well as native speakers of the target
language. Like the Canadian immersion students, the El Paso BIPs are in effect
isolated from peer groups which are monolingual in the target language, and
while the Canadian analogy begins to break down at this point at least partially,
the linguistic consequences are similar. The chief difference between the two
situations is that for reasons based as much on class as on ethnicity, the aver-
age El Paso hispanomonomatriphonic child typically lacks access to social net-
works that are monolingual in English, while the typical Anglo-Canadian does
not take advantage of the access putatively available. In El Paso, "neighbor-
hood" is the chief determinant of language choice (see Teschner 1981): some
neighborhoods--not surprisingly those where hispanophones enjoyed primacy of
settlement, or where Hispanics constitute almost a totality of the population- -
are simply known as "Spanish-speaking," and social pressures insure that Hispan-
19
ics from outside those neighborhoods speak Spanish when visiting them, whether
Spanish is or is not the outsiders' preferred or dominant language. When en-
gaged in commercial or social activity outside the Spanish-bastion areas, their
residents readily find fellow Hispanics to interact with, and in Spanish. This
is not to make the claim that in. El Paso, all intra-Hispanic interaction takes
place exclusively in Spanish, or that the majority of El Paso Hispanics are
monolingual in that language. On the contrary, shift to English is taking
place; yet immigration from Mexico is also proceeding apace. What character-
izes El Paso then is a steady-state bilingualism manifesting a slow shift to
English dominance and/or preference (though almost never English monolingual-
ism), alongside a steady influx of monolingual hispanophones and a constant
level of overall societal Spanish dominance and/or preference. In effect, for
every Hispanic who shifts to English dominance and/or preference, a monolingual
hispanophone is added to the population, particularly in.certain neighborhoods,
which brings us back to the point at issue: the degree to which an immersion
program or for that matter any program can provide the child of certain- neigh-
borhoods, certain backgroundswith the meaningful exposure that he or she needs
to achieve parity on English-medium test scores--not to mention (following Edel-
sky) all those far-more relevant activities inside school and out--with those
"district-wide" students whose home-language circumstances or neighborhoods
of residence have enabled them to bring proficient English to school from the
outset, whether they also speak Spanish or not.
I have long maintained (see Teschner 198C, 1981, 1982, 1983, Parisi and
Teschner 1984) that, paradoxically, in an area where bilingualism is both
profuse and steady-state, i.e., not accompanied by wide-scale shift, individ-
ual acquisition of the area's second language is typically problematic: be-
cause there is always someone around who speaks your primary language, incen-
tives to become proficient in the other tongue are correspondingly reduced.
Not that for our children in BIP and TBE programs these incentives are lack-
20
20
ing altogether. A 1982 UTEP masters thesis survey (Hammersmith 1982) shows
that what amounts to a totality of parents of children in El Paso's bilingual
programs--TBE only at the time--want their offspring to learn English, and
fully expect that the schools will achieve that goal.. Yet--the survey re-
veals--few if any of these parents speak.English at home or expect that their
children will become anglicized; indeed, they are confident that their child-
ren will remain as loyal to language, class and culture as they themselves
have been. Once again we are in Canada, mutatis mutandis, save for the not-
inconsiderable differences in social class between the two populations. Which
brings us una vez encore to the central question, restated somewhat in light
of what I have just discussed: can the task of producing proficient bi-
linguals be a school task alone? Obviously this question is rhetorical,
though only in part In some El Paso neighborhoods, the products of BIP and
even TBE will experience integration--eventually or even rapidly--into a peer
population which is English- preferent, -dominant, or -monolingual. In other
neighborhoods, the products of bilingual schooling are likely to remain
segregated or largely so, either because they live in public housing projects
(always prone to ostracism and self-ostracization), or because the neighborhood
is both historically and statistically Spanish-monolingual or nearly so.
The quote that appears in the title of the present paper has been taken
from Housen and Baetens Beardsmore (1987:87). By placing such a caveat in such
a prominent position I do not mean to imply that Cummins is unaware of the impor-
tance of adequate exposure to the L2 in school and environment, along with suf-
ficient motivation to learn it Such is not the case. In any early explanation
of his developmental interdependence hypothesis, for example, Cummins explicitly
cites environmental sufficiency as a necessary variable (" . . instruction is
effective in promoting CALP which will manifest itself in both languages given
adequate motivation and exposure to both languages either in school or wider
environment . . . Thus, the relationships . . . presuppose motivational in-
1)4 1.044
21
volvement and adequate exposure to Ll and/or L2." (1979a204),. Similar
sentiments appear in Cummins 1980a:88 and 90, 1980b:185, 1984b: -100 ("For
example, minority children' are unlikely to perforili adequately in English if
exposure to English has been minimal")', 1984b:265 -(" . . . students instructed4
for much of the day through a minority language Will perform at least as well
in English academic skills as equivalent students instructed totally through
English, given adequate motivation to learn English and sufficient exposure
to the language either within or,outside school"), and 1984c:33. The problem
is that Cummins not Infrequently makes assumptions elsewhere which teaOly lend
themselves to misinterpretation or what appears to be an outright dismissal
of the environmental/motivational caveat, sometimes by Cummins himself, as
quotes from Cummins earlier in this paper have shown (1979c:243, 1981f:44).
Thus the explicit assumption in 1980b:185 ("Because the majority language is
the language of the streets and of T.V. there is usually no lack of exposure
or motivation to acquire it"), the patently site-specific conclusion which
explains away the findings of a Toronto study ("Bhatnager (1980) reports that
immigrant students who used Ll exclusively with parents and siblings also per-
formed significantly worse than those who used both Ll and L2. However, it
seems likely that this finding can be attributed to the fact that only those
students who had immigrated relatively recently would use Ll exclusively" (1981f:
33, emphasis added)), the assumption contained in 1984b:158 (" . . . the appro-
priate implication from the Canadian immersion data is that the language whose
development is most likely to-be neglected (i.e. the minority or subordinate
language) should be strongly promoted in the school programme in order to pro-
duce an additive form of bilingualism" (emphasis added)), and the unexamined
assumptions made by McLaughlin 1984-85 and restated uncritically by Wesche
(1987:76) in her review of McLaughlin, who, "in the context of his discussion
of why Canadian immersion models are not appropriate for minority language
children . . . proposes 'reverse immersion' for children who come to school
22.
22
without English skills. In this model the weaker (first) language would be
taught almost exclusively durinOhe primary years, while oral second (English)
language skills would initially be absorbed on the playground and in activities
requiring little English proficiency . . . " (How is it possible to speak of
Ll as the weaker language in children who come to s:hool without any L2?)
The chief problem with Cummins here is an entirely understandable Canado-
centrism, which has prompted him on more than one occasion to forget the advice
that he himself had given in an early publication (Cummins and Gulutsan 1974:
261): " . . . each bilingual education project must be planned according to
the needs of the particular area in which it is located . . . " Thus what is
valid for 'a° presumably bilingual-by-first-grade groups of francophone students
in Winnipeg, an English-dominant city in a markedly English-dominant region
(H4bert 1976, as cited in Cummins 1979c and elsewhere) cannot be assumed as
valid for all alloglots everywhere, and is demonstrably not valid for the typi-
cal LEP or NEP hispanophone El Paso six year old. As Wong-Fillmore has noted,
one cannot take "street" acquisition for granted: "'Contrary to the usual
assumption that children learn language mainly from peers outside the classroom
and not from teachers, it appears that for many LEP students the only place in
which they come into regular contact with English speakers is at school. Thus,
language learning, if it is going to take place at all, is going to have to
happen at school. (p. 19 {of Gass and Madden 1988})"(Scarcella and Perkins
1987:350). Wong-Fillmore leaves unanswered the degree to which "playground"
acquisition can be-predicted, as well she might, given the vast host of varia-
bles attendant thereunto: ratio of Lx-monolingual to Ly-monolingual children
and degree and stability of bilingualism; nature of the relationship between
speakers of Lx and Ly in the narrower and wider communities; stability of Lx/
Ly settlement in that community (i.e., how rapidly is the neighborhood changing?);
expectations regarding the direction of language shift if any; social class and
its relationship to expected patterns of communication--in short, the entire
23
23
gamut of topics investigated by ethnographers and sociolinguists.
In school/community settings typified by considerable degrees of isolation,
the magic of immersion education is hard-pressed to work--to the extent that
the immersion classroom alone can be counted on to produce much more than a
mere approximation of the target language (see for example Hammerly 1987 and
my own discussion below). This paper has sounded this theme before and will
now examine it in greater depth.
Until recently, Cummins' own writings took the unqualified or near-
unqualified success of Canadian immersion education (CIE) as a given, to be
referred to in passing as an example of the easy feasibility of home-school
language switch at least for majority-language children (see for example
1984b:13). His first admission that Canadian anglomonomatriphones' command
of French might be less than had been thought came in 1983, when he admits that
"In terms of French achievement, by grade 6 students' oral and written recep-
tive skills are close to those of native speakers. However, although they
are fluent, their productive skills are not native-like." (1983a:375) It was
not until 1986 that Cummins (writing with Merrill Swain) fully credited the
degree to which adequate interactive exposure t,5-the L2 is necessary for the
alloglot to become proficient in it. In a discussion of differences between
native speakers and CIE students, Cummins and Swain write that these
. . .were highly significant on most grammatical measures and on
those discourse and sociolinguistic measures where grammatical know-
ledge was essential for the production of correct linguistic forms.
Additionally, it has been found that early immersion students approach
native-speaker levels of proficiency in French reading and listening
measures {i.e., in receptive skills} by the end of elementary school
but significant differences remain with respect to oral and written
{i.e., productive} grammatical skills . . . " (1986:209)
Why, the authors ask, do "immigrant students attain proficient L2 grammatical
24
24
skills within a relatively short time whereas French immersion students continue
to experience diffitulty in these areas?" (Swain and Cummins 1986:211) The
answer is a logical extension of their findings anent exposure: "The same hypo-
these appear relevant to this phenomenon. Immigrant students tend to-gain
peer-appropriate grammatical skill's, at least in the oral modality, considerably
more rapidly than is the case for immersion students precisely because their
contact with and use of the L2 is far greater." (ibid.) In the El Paso context,
"provided that" rather than "because" would appear to be the more appropriate
conjunctive element--"'provided that' their contact with and use of the L2 is
far greater (or sufficient)."
Recent publications by Canadians give the distinct impression that CIE is
undergoing considerable reevaluation. Though Swain and Lapkin (1986) continue
to emphasize the positive aspects of CIE even as they note the negative, a half-
dozen other researchers have come down increasingly hard on CIE for producing
just the sort of graduates who "continue to experience difficulty" in French
production after thousands and thousands of hours of classroom exposure; thus
Adiv (1980), Lister (1987), Pawley (1985), Pellerin and Hammerly (1986), Singh
(1986), and in particular Hammerly (1987), who continued Spilka (1976)'s pion-
eering examination of the oral production of French immersion students. Ham-
merly goes so far as to state that CIE is "linguistically . . . a failure" and
that it reveals."irremediable problems" (Hammerly 1987:399).
It is clear from the works just cited as well as (more importantly for
our immediate purposes) Cummins and Swain themselves that classroom settings
alone cannot be expected to produce BICS (Basic Interpersonal Communicative
Skills) though they are capable of developing CALP (Cognitive Academic Language
Proficiency). I have deliberately employed these dichotomizing acronyms, which
Cummins himself has now abandoned (in favor of "context-embedded" and "context-
reduced"--see 1983b:125 et alibi), so as to draw attention to the changes in
the way that Cummins views the acquisition of the ability to talk with peers,
25
25
among other activities (as opposed to perfor. On tests, etc.). Whereas the
early CUMmins gave the impression that BICS essentially developed by, itself,
given proper exposure and motivation, by 1980 CumMins states that " . . . in an
L2 context, where proficiency is as yet inadequately developed, a wider range
of interpersonal communicative tasks are cognitively demanding than in an Ll
context" (1980c:31-32), and three years later is cited (in John 011er's intro-
duction to Cummins' contribution to the former's volume on testing) as explic-
itly recognizing that "face-to-face activities can be cognitively demanding,
i.e., require inference etc.' (Cummins, personal communication . . . )." (1983b:
108) Under conditions of social isolation from monomatriphonic speakers of the
majority language, "BICS" acquisition may take longer than the time Cummins
has posited, on the basis of his Toronto findings, as Wald (1984:61) points
out in a commentary on Cummins' hypotheses:
The claim that second language fluency is generally achieved in
one and a half to two years does not correspond to the behavioral
characteristics observed for 10-12 year old Mexican American children
in the Los Angeles area. Quite generally, it has been found that
students of less than four to five years residence in an English-
speaking {sic?} environment preferred Spanish in spontaneous con-
versation, and have difficulty in competing for the floor in English
with age -mates of earlier ages of arrival . . .
As we know from Swain (1983 {as reported on by Housen and Baetens Beards-
more 1987:89 }, from whom I quote at length as befits the importance of the
commentary, "comprehensible input, or the i + 1 hypothesis alone, is insuffi-
cient for the acquisition of high levels of L2 proficiency, based on results
obtained on Canadian grade 6 immersion pupils," who had--using Cummins' early
terminology--"acquired CALP" quite nicely but lacked important aspects of BICS
(again to oversimplify).
These subjects performed satisfactorily on subject matter tests'and
LI4c.
26
consequently must have understood what was taught through the L2,
which suggests that it is not input per se that is important to L2
acquisition, but input that occurs in interaction where meaning is
negotiated. With immersion pupils, input is derived mainly from
listening to teacher talk, so the less than native-like grammatical
competence can only be accounted for by the inadequacy of the input
hypothesis. On the other hand, Swain claims that output fulfills
a vital role in the process of L2 acquisition in that it enables
the acquirer to apply the available linguistic resources in a mean-
ingful way. This pushes the acquirer toward the delivery of a mes-
sage that is conveyed as precisely, coherently, and appropriately
as possible, enabling experimentation with target language structures
by trial and error. Further, output forces the acquirer to pay equal
-atthntion to-semantic and syntactic features, whereas the decoding of
input allows for meaning to be derived from a combination of linguis-
tic and extra-linguistic cues without any necessary focus on struc-
ture. {Immersion students} are limited in output because the wider
environment does not sufficiently =push the speaker ir. the,active use
of the L2. Nor do immersion children-get much opportunity to inter-
act with native-speaker peers of the target language.
While ambiental- conditions among BIP students in El Paso and similar settings do
not constitute exact replicas of conditions .1 Canada as these inform the con-
clusions of Swain and others, an analogy can nonetheless be made. In the BIP
El Paso context one must speak of wider environments, plural: on the one hand
a wider environment of family, friends and neighbors with whom the Ll (Spanish)
is patently the expected medium of communication, and on the other hand a wider
environment that consists of both the immediate expectations of the school sys-
tem and the future expectations of the wider society - -these latter however dil-
uted by the obvious presence of Spanish' as in some respects a co-equal "public"
27
27
language for significant numbers of adult Hispanics in places like El Paso and
elsewhere. The point is that though environmental pressures toward English are
certainly stronger in the BIP El Paso context than they are toward French among
immersion students in, say, Vancouver, Edmonton or Winnipeg, they are not as
strong as are similar pressures toward English among the immigrant students of
Toronto with whom Cummins is most familiar. As is so often the case in socio-
linguistics, we are dealing not with a dichotomy but with a continuum.
At this point in my analysis the temptation is strong to folio.; intuition
("common sense") and recommend that BIP--let alone TBE--be replaced by what
Cummins has described as monolingual majority-language immersion programs "which
take into account the need to provide minority students with modified second-
language input but which dispense with bilingual teachers and first-language
literacy promotion." (1983a:379) While -I do not deny that the empirical find-
ings form such a program in El Paso would be of considerable interest to research-
ers, nor the possibility that an MIP (monolingual immersion program) might prove
a not-unattractive educational alternative under cer:ain circumstances, I never-
theless agree with Cummins' conclusion that for some types of alloglottic grade-
school populations, an Ll component is a critical necessity. Cummins advocates
the inclusion of Ll in the curriculum on two grounds: first, that the devel-
opmental interdependence and the threshold hypotheses predict that cognitive
grWti in L2 is-contingent upon continued cognitive development in Ll (and vice
versa), and, second, that "An examination of global trends in the educational
achievement of minority language children suggest that the groups who perform
poorly in school are those who have been discriminated against economically
and educationally and, as a consequence, are characterized by insecurity in
relation to their own language and culture and ambivalence towards the dominant
language and culture of the society . . . (1984a:64) Of the two arguments I
find the second the far more compelling, not only because of the doubts that
Martin-Jones and Romaine (1986) and others have cast on the feasibility of
28
28
subjecting a developmental interdependence hypothesis to empirical examination
as a single variable, but--and in larger part--because of what Cummins himself
has had to say on the utility of the distinction between additive and subtrac-
tive bilingualism as interrelated with certain, variables among minority language
populations; thus while (1984a:64)
The distinction between additive and subtractive bilingualism is
useful in delineating a- gross difference between bilingual acquisi-
tion contexts . . . the distinction fails to-explain the very large
differences in educational- performance between groups of minority
language children in subtractive situations. Not all groups of
minority language children perform poorly at school; in fact, studies
carried out by the Toronto Board of Education .- . show that, with
the exception of children whose mother tongue is French, minority
language children who were born in Canada tended to have higher
academic achievement and were more likely to be in high academic
streams than unilingual children, born in Canada, whose mother tongue
was English. ... However, for groups of minority children such as
the French in Ontario . . . , the Finns in Sweden, and Hispanic and
Indian children in North America, substantial deficits in educational
achievement have been reported. These findings cannot be wholly
accounted for by the notion of subtractive bilingualism since the
minority language children in the Toronto Board of Education studies
are also in subtractive situations.
So there are minority language children and minority language children, and in
this sense Cummins makes thoughtful use of John Ogbu's tripartite classification
of minority groups into "autonomous," "immigrant" and "caste" minorities (Cummins
1984e:82): though the last of these may indeed be mostly immigrant in origin, as
is the case for example with Mexican-Americans, they are best classified as a
"caste" minority for reasons social and historical. What Cummins means by
29
"ambivalence towards the dominant language" is--in pop psych terms--that the
alloglottic caste minoritarian is involved in a like/dislike, or love/hate, or
push/pull relationship with the particular L2: it is the widely-accepted lang-
uage necessary for economic advancement but it is also the language of the his-
torical enemy, etc. The world abounds in such situations and they need not be
delved into further here. Suffice it to say that in a program such as BIP, one
can envision the presence as authority figure (teacher) of a fellow allophonic
though bilingual ethnic as serving to validate the presence of the ambivalently-
viewed L2 as the primary medium of instruction, all the more so since this same
authority figure also validates the heretofore insecurely-viewed Ll by using
it as a medium of instruction as well.
Attitudinal considerations, then, convince me that the Ll component is
necessary to the EPISD's immersion program for hispanomonomatriphonic students
and under similar conditions in-the-United States and elsewhere, just as the
relative superiority of the Bilingual Immersion Program's test results has con-
vinced me that in settings such as El Paso where the minority language enjoys
considerable strength and salience, some form of majority-language immersion is
preferable both to transitional bilingual education and minority-language immer-
sion, which do not provide an adequate basis for accessing a language that is
socially remote to greater or lesser degrees, and, thus, for the inevitable
transition to instruction through the majority language exclusively. My ad-
vocacy of BIP, however, should not be construed as an insistence that this pro-
gram as presently constituted represents the best of all possible schooling
worlds. Instead, BIP represents a superior educational alternative, a step on
the road to a better education, but not a panacea.
30
NOTES
1Between 1974 and ca. 1980, Cummins' publications almost always list his
first name as James, but since then it has usually appeared as Jim. The present
paper's Bibliography has not sought to follow these shifts.
2Torrance et al. 1982's statistical analysis and testing program are mini-
mal at best and could not serve as an adequate basis for comparison with the
EPISD's or any other district's program. I am obligated to Tim Allen, Director
{though not until after 1982}, Second Language Education, San Diego City Schools,
for sending me a copy of Torrance et al. 1982 and discussing various curricular
matters with me over the phone.
'3My thanks go to Ken Thomas, Associate Superintendent for Curriculum, EPISD,
and the staff members of the Office for Research and Evaluation, EPISD. for pro -'
viding me with a copy of the report.
-31
REFERENCES
Adiv, Ellen. 1980. An analysis of second language performance in two types of
immersion programs. Bulletin of the Canadian Association of Applied Ling-
uistics, 2, 139-152.
Bhatnager, Joti. 1980. Linguistic behavior and adjustment of immigrant
children in French and English schools in Montreal. International Jour-
nal of Applied Psychology, 29, 141-158.
Cummins, James and M. Gulatsan. 1974: Bilingual education and cognition.
Alberta Journal of Educational Research, 20, 259-269.
Cummins, James. 1977a. Alberta program promotes minority languages. The
Linguistic Reporter, 20(2), 3, (November 1977).
1977b. Cognitive factors associated with the attainment of
intermediate levels of bilingual- skills. Modern Language Journal, 61,
3-12.
1977c. Delaying native language reading instruction in immersion
programs: A cautionary note. Canadian Modern Language Review,- 34,- 46 -49.
1977d. Psycholinguistic evidence. In Bilingual education:
Current perspectives (Vol. IV, "Education"), 78-89.
1978a. Bilingualism and educational development in anglophone
and minority francophone groups in Canada. Interchange on Educational
Policy, 9(4), 40-51.
1978b. Bilingualism and the development of metalinguistic aware-
ness. Journal of Cross-Cultural Psychology, 9, 131-149. (Also appears in
Cummins and Swain 1986:20-32.)
1978c. The cognitive development of children in immersion
programs. Canadian Modern Language Review, 34, 855-883.
1978d. Educational implications of mother tongue maintenance
in minority-language groups. Canadian Modern Language Review, 34, 395-416.
Rs 2
32
1978e. Immersion programs: The Irish experience. Interna-,
tional Review of Education, 24, 273-282.
1978f. Language and children's ability to evaluate contradic-
tions and tautologies: A critique of Osherson and Markman's findings.
Child Development, 49, 895-897.
1978g. Metalinguistic development of children in bilingual
education programs: Data from Irish and Canadian Ukrainian-English programs.
In Michel Paradis (ed.), The Fourth LACUS Forum. Columbia, South Carolina:
Hornbeam Press. Pp. 29-40.
and Robert Mulcahy. 1978. Orientation to language in Ukrainian-
English bilingual children. Child Development, 49, 1239-1242.
1979a. Cognitive/academic language proficiency, linguistic
interdependence, the optimum age question and some other matters. Working
Papers on Bilingualism, 19, 197-205.
1979b. The language and culture issue in the education of
minority language children. Interchange on Educational Policy, 10(4),
72-88.
1979c. Linguistic interdependence and the educational develop-
ment of bilingual children. Review of Educational Research, 49, 222-251.
(Also in Bilingual Education Paper Series, Vol. 3, No. 2. Los Angeles,
California: National Dissemination and Assessment Center/California State
University at Los Angeles, 1979. 63 pp.)
1980a. The construct of language proficiency in bilingual
education. GURT {Georgetown University Roundtable on Languages and Ling-
uistics}, 1980, 81-103.
1980b. The cross-lingual dimensions of language proficiency -:
Implications for language education and the optimal age issue. TESOL
Quarterly, 14, 175-187.
1980c. The exit and entry fallacy in bilingual education.
33
33
NABE Journal, 4, 25-60.
1980d. Psychological assessment of immigrant children: Logic
or intuition. Journal of Multilingual and Multicultural Development,
1, 97-111.
1981a. Age of arrival and immigrant second language learning
in Canada: A reassessment. Applied Linguistics, 2, 132-149.
1981b. Bilingualism and minority language children. Toronto:
Ontario Institute for Studies in Education (OISE).
1981c. Educational success for Canadianrminority language
children: The role of mother tongue development. Canadian Journal of
Italian Studies, 4(3-4), 299-315 (Spring-Summer 1981).
1981d. Empirical and theoretical underpinning of bilingual
education. Journal of Education, 163(1), 16-29 (Winter 1981).
1981e. Four misconceptions about language proficiency in
bilingual education. NABE Journal, 5(3), 31-45 (Spring 1981).
1981f. The role of primary language development in promoting
educational success for language minority students. In California State
Department of Education, Schooling and language minority students: A
theoretical framework. Los Angeles: Evaluation, Assessment and Dissem-
ination Center/California State University at Los Angeles. Pp. 5-49.
1983a. Bilingualism and special education: Program and
pedagogical issues. Learning Dtiability_Quarterly, 6,- 373 -386.
1983b. Language proficiency and academic achievement. -In
John W. Oiler, Jr. (ed.), Issues in language testing research. Rowley,
Massachusetts: Newbury House Publishers. Pp. 108-129. (Also in. Cummins
and Swain 1986:138-161.)
1983c. Language- proficiency, biliteracy and French immersion.
Canadian Journal of Education, 8, 117-138.
1983d. Mother tongue development as educational enrichment.
34
34
In Heritage language education. Toronto: Ontario Ministry of Education/
Ontario Institute for Studies in Education (OISE). -Pp. 40-43.
and Merrill Swain. 1983. Analysis-by-rhetoric: Reading the
text or the reader's own projections? A reply to Edelsky et al. Applied
Linguistics, 4, 23-41.
1984a. Bilingualism and cognitive functioning. In Stan Shapson
and Vincent D'Oyley (eds.), Bilingual and multicultural education: Canadian
perspectives. Clevedon, Avon, UK: 'Multilingual Matters Ltd. Pp. 55-67.
-1984b. Bilingualism and special education: Issues in assessment
and pedagogy. San Diego, California: College Hill Press.
1984c. Implications of bilingual proficiency for the education
of minority language students. In Patrick -Allen and Merrill Swain (eds.),
Language issues and education policies: Exploring Canada's multilingual
resources. Oxford, UK: Pergamon Press. Pp. 21-34.
1984d. Linguistic minorities and multicultural policy in Canada.
In John Edwards (ed.), Linguistic minorities, policies and pluralism.
London, UK: Academic Prest. Pp. 81-105.
1984e. The minority language child. In Stan Shapson and
Vincent D'Oyley (eds.), Bilingual and multicultural education: Canadian
perspectives. Clevedon, Avon, UK: Multilingual Matters Ltd. Pp. 71-92.
.1984f. Minority students and learning difficulties: Issues in
assessment and placement. Neurolinguistics, 13, 47-68. (Also in Michel
Paradis and Yvan Lebrun (eds.), Early bilingualism and child development.
Lisse: Swets and Zeitlinger, 1984. Pp. 47-68. Also--in revised form- -
in Cummins and Swain 1986:183-204.)
1984g. Wanted: A theoretical framework for relating language
proficiency to academic achievement among bilingual students. In Charlene
Rivera (ed.), Language proficiency and academic achievement. Washington DC:
Center for Applied Linguistics/Clevedon, Avon, UK: Multilingual Matters
35
Ltd-. Pp'. 2-19.
1984h. Language_proficiency and academic achievement revisited:
A- response. In Charlene-Rivera (ed.), Language proficiency and academic
achieveMent. Washington DC: Center for Applied Linguistics/Clevedon,
Avon, UK: Multilingual Matters Ltd. Pp. 71-76.
et al. 1984. Linguistic interdependence among Japanese and
Vietnamese immigrant students. In Charlene Rivera (ed.), Cummunicative
competence approaches to language proficiency assessment: Research and
application. Clevedon, Avon, UK: Multilingual Matters Ltd. Pp. 60-81:
1986a. Empowering minority students: A framework for inter-
vention. Harvard Educational Review, 56(1), 18-36 (February 1986).
1986b. Psychological assessment of minority students:. Out
of context, out of lbcus, out of control? Journal of Reading, Writing,
and Learning Disabilities International, 2, 9-19:
and Merrill Swain. 1986. Bilingualism in education: Aspects
of theory, research and practice. London Lind-New York: Longman.
Edelsky, Carole et al. 1983. Semilingualism and language deficit. Applied Ling-
uistics, 4, 1-22.
Edelsky, Carole-. 1984. The effect of 'theory' on theory--And other phenomena.
SouthWest_Journal of Linguistics, 7, 74-86,
Gass, Susan and-Charles Madden (eds.).. 1985. Input_in_second_languageacquisi.!
tion. Rowley, Massadhusetts: Newbury House Publishers.
Glazer,- Nathan and-Jim Cummins. 1985. Viewpoints on bilingual education. Equity
and Choice, 2(1), 47-52 (Fall 1986).
Hammerly, Hector: 1987. The immersion approach: Litmus test of second language
acquisition through classroom communication. Modern Language Journal, 71,
395-401-
'Hammersmith, Ann M. 1982. Assessing the acquisition of English-as a second
language in elementary school: An El Paso study. thesis, University
If
36
of Texas at El Paso.
Hebert, R. 1976. Rendement academique et langue d'enseignement chez les
eaves franco-manitobains. Saint-Boniface, Manitoba: Centre de recherches
du College Universitaire de Saint-Boniface.
Housen, Alex and Hugo Baetens Beardsmore. 1987. Curricular and extra-curricular
factors in multilingual education. Studies in Second Language Acquisition,
9, 83-102.
Interim Report 1987. Interim report of the, five-year bilingual education pilot,
E1 Paso Independent School District, 1986-87 school year. El Paso, Texas:
El Paso Independent School District, Office for Research and Evaluation.
Lambert, Wallace E. and G. Richard Tucker. 1972. Bilingual education of
children: The St. Lambert experiment. Rowley, Massachusetts: Newbury
House Publishers.
Lister, Roy. 1987. Speaking immersion. Canadian Modern_ Language Review, 43,
701-717.
Martin-Jones, Marilyn and Suzanne Romaine. 1986. Semilingual ism: A half-baked
theory of communicative competence. Applied Linguistics, 7, 26-38.
McLaughlin, Barry. 1984-85. Second language acquisition in childhood. 2 vols.
2nd ed. Hillsdale, New Jersey: Erlbaum.
Pa' :si, Gino and Richard V. Teschner. 1984. Unitary placement testing for
Spanish: Can a single instrument adequately serve native speakers and
non-natives alike? Foreign Language Annals, 17, 173-184.
Pawley, Catherine. 1985. How bilingual are French immersion students?
Canadian Modern Language Review, 41, 865-876.
Pellerin, Micheline and Hector Hammerly. 1986. L'expression orale apres treize
ans d' immersion francaise. Canadian Modern Language Review, 42, 592-606.
Scarcella, Robin and Leroy Perkins. 1987. Review article: Shifting gear's:
Krashen's input hypothesis. Studies in Second Language Acquisition, 9, 347-354.
Singh, Rajendra. 1986. Immersion: Problems and principles. Canadian Modern
7
Language Review, 42, 559-571.
Spilka, I4ne. 1976. Assessment of second-language performance in immersion
programs. Canadian Modern Language Review, 32, 543-561.
Swain, Merrill. 1983. Communicative competence: Some roles of comprehensible
input and comprehensible output in its development. Paper presented at the
Tenth Michigan Conference on Applied Linguistics, University of Michigan,
Ann Arbor.
and James Cummins. 1982. Bilingualism, cognitive functioning
and education. In Valerie Kinsella (ed.), Surveys 1: Eight state-of-the-
art articles on key areas in language teaching. Cambridge, UK: Cambridge
University Press. Pp. 23-37. (Also in Language Teaching and Linguistics:
Abstracts, 4-18 (1979), and in Cummins and Swain 1986:7-19.)
and James Cummins. 1986. Towards a theory of bilingual pro-
ficiency development. In Cummins and Swain 1986:207-213.
and Sharon Lapkin. 1986. Immersion French in secondary schools:
'The goods' and 'the bads'. Contact, 5(3), 2-9.
Teschner, Richard V. 1980. Spanish as a Texas language: Are things beginning
to Leuven up? Paper presented at the annual meeting of the Modern Language
Association of America (MLA), Houston, Texas.
. 1981. Historical-psychological investigations as complements
to sociolinguistic studies in relational bilingualism: Two Mexican-American
cases. The Bilingual Review/La Revista Bilingde, 8, 42-55.
. 1982. Second-lan ua e ac uisition and forei n lan ua e teachin
Spanish language programs at a university on the U.S.-Mexican border. In
Florence Barkin et al. (eds.), Bilingualism and language contact: Spanish,
English, and Native American languages. New York: Teachers College Press.
Pp. 228-240.
. 1983. Spanish Oacement for native speakers, nonnative speakers,
and others. ADFL Bulletin, 14(3), 37-42 (March 1983).
38
38
Texas Education Agency. 1982. Texas Education Agency Bulletin, 9 (October),
451-464.6.
Texas Education Code. 1982. University of Texas Law School Bulletin 1982,
no. 7. Pp.163-168.
Torrance, Dianne W. (ed..) et al. 1982. An exemplary approach to bilingual
education: A comprehensive handbook for implementing an"elementary-level
Spanish-English language immersion program. San Diego, California: San
Diego Unified School District/ESEA Title VII Bilingual Demonstration Pro-
ject, Publication #I- B- 82 -58.
UNESCO. 1953. The use of vernacular languages in education. (Monographs
on fundamental education, 1953.) Paris: UNESCO.
Wald, Benji. 1984. A sociolinguistic perspective on Cummins' current frame-
work for relating language proficiency to academic achievement. In Char-
lene Rivera (ed.), Language proficiency and academic achievement. Washing-
ton DC: Center for Applied Linguistics/Clevedon, Avon, UK: Multilingual
Matters Ltd. Pp. 55-70.
Wardhaugh, Ronald. 1983. Language & nationhood: The Canadian experience.
Vancouver, British Columbia: New Star Books.
Wesche, Marjorie B. Review of McLaughlin 1984-85 (q.v. supra). Modern Language
Journal, 71, 74-76 (1987).
39